Today, Fotmob launched Premier League physical performance data (Figure 1). At first glance, it is somewhat disappointing; these metrics will inevitably become a new tool for some people to 'compare intensity' and bring a new kind of enjoyment to 'data football fans.' It must be pointed out here: judging intensity solely based on distance is a rather poor approach. Moreover, we need to be cautious when using Fotmob's data, because there are many issues behind it. Here are three points to note: 1. Data source is extracted from broadcast footage. In the field of computer vision, after years of rapid development, there are already many mature demos for object detection and tracking, and the entry threshold is not high (web link). Those interested can study it themselves and are welcome to discuss it; it will not be elaborated here. However, the problem is that broadcast angles are usually limited by the camera range, field size, and number of people in the frame. The resolution of the footage and occlusion between players can also greatly affect the accuracy of recognition, so the precision of the extracted data is very limited. 2. Only provides season summaries, lacking context. Currently, Fotmob only provides 'total running distance per season.' This is coarse-grained data and cannot reveal a player's specific output in different match stages or under different role tasks, and it cannot describe the explosive runs required in key tactical moments. As repeatedly pointed out in past research and discussions, the physical output of different players during defensive and offensive phases differs greatly, and its tactical significance is far beyond what 'total distance' can capture. 3. No accurate playing time and effective time. Fotmob does not consider stoppage time, let alone effective playing time. However, in load analysis, even a 5% difference in meters per minute can imply a completely different injury risk. Additionally, total distance over a season is positively correlated with total time played in the season. Playing every match fully naturally increases the distance, but it does not reflect the physical reserves required for a single game. Therefore, ignoring playing time and comparing players' physical output solely based on 'total distance covered in a season' may produce conclusions that seem reasonable, but this naive method of comparison is inherently flawed. Taking the 24/25 Champions League season as an example, Florentino of Benfica and Vitinha of PSG, both defensive midfielders in a 4-3-3 system, each ran a total distance of 12,966 meters in a single match against Atlético—what a coincidence! But if we look a little closer: Florentino (Benfica vs. Atlético de Madrid) Total distance: 12,966m Match time: 98 minutes 30 seconds M/Min: 131.55 m/min Vitinha (PSG vs. Atlético de Madrid) Total distance: 12,966m Match time: 93 minutes 54 seconds M/Min: 138.00 m/min The difference in m/min between the two is already 5%, which corresponds to completely different match contexts. Figure 2 also shows the players’ running patterns, especially in high-speed running and sprints, which differ considerably. This simple example also demonstrates that without context, even the premise of comparing 'who ran more' does not hold. I strongly support open data. After all, football data analysis has fallen into the storytelling trap for more than a decade. Only by liberating the valuable data from clubs can we further connect real match scenarios at both macro and micro levels and uncover true value. Compared to event data, running distance is closer to the 'exercise' itself and better reflects 'what the player's body is actually experiencing.' But once taken out of context and structure, it loses explanatory power. Essentially, total running distance is an indicator that can be used to explain anything, yet at the same time, it explains nothing. A true data fan should have both the technical foundation for independent data analysis and the ability to understand the game. Only in this way can data be interpreted in real-game scenarios. Otherwise, no matter how much hearsay statistics you have or how much data you store, if you don’t look at the game firsthand, these numbers are ultimately just decoration and will only mislead the analysis. Instead of debating who ran more or who had higher intensity based on a summarized dataset with unclear premises, we should be more concerned with whether a particular movement completed a tactical task, whether it changed the situation, what kind of physical consumption or damage it caused, and whether the movement was worthwhile. These are the questions that physical performance analysis should answer.